
import pandas as pd
import numpy as np
from matplotlib import pyplot as plt
%matplotlib inline
import seaborn as sns
from sklearn.preprocessing import LabelEncoder
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import cross_val_score
from sklearn.ensemble import RandomForestRegressor, GradientBoostingRegressor, BaggingRegressor
from sklearn.svm import SVR
from sklearn.tree import DecisionTreeRegressor
from sklearn.metrics import r2_score
from xgboost import XGBRegressor
from pandas_profiling import ProfileReport
from collections import Counter
import math
import statsmodels.api as sm
from sklearn import metrics
from sklearn.preprocessing import RobustScaler
from sklearn.model_selection import KFold, cross_val_score
from sklearn.metrics import mean_squared_error , make_scorer
from sklearn.linear_model import ElasticNetCV, LassoCV, RidgeCV
from sklearn.pipeline import make_pipeline
from sklearn.linear_model import LinearRegression
from sklearn.linear_model import LinearRegression
import warnings
warnings.filterwarnings('ignore')


Salestrain=pd.read_csv("Train.csv")


Salestrain


# Checking the dimensions of the training dataset
Salestrain.shape

(8523, 12)


Salestrain.head(10)


Salestrain.tail(10)


ProfileReport(Salestrain)

Summarize dataset:   0%|          | 0/5 [00:00<?, ?it/s]

Generate report structure:   0%|          | 0/1 [00:00<?, ?it/s]

Render HTML:   0%|          | 0/1 [00:00<?, ?it/s]


Salestrain.describe()


Salestrain.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 8523 entries, 0 to 8522
Data columns (total 12 columns):
 #   Column                     Non-Null Count  Dtype  
---  ------                     --------------  -----  
 0   Item_Identifier            8523 non-null   object 
 1   Item_Weight                7060 non-null   float64
 2   Item_Fat_Content           8523 non-null   object 
 3   Item_Visibility            8523 non-null   float64
 4   Item_Type                  8523 non-null   object 
 5   Item_MRP                   8523 non-null   float64
 6   Outlet_Identifier          8523 non-null   object 
 7   Outlet_Establishment_Year  8523 non-null   int64  
 8   Outlet_Size                6113 non-null   object 
 9   Outlet_Location_Type       8523 non-null   object 
 10  Outlet_Type                8523 non-null   object 
 11  Item_Outlet_Sales          8523 non-null   float64
dtypes: float64(4), int64(1), object(7)
memory usage: 799.2+ KB


Salestrain.isna().sum()

Item_Identifier                 0
Item_Weight                  1463
Item_Fat_Content                0
Item_Visibility                 0
Item_Type                       0
Item_MRP                        0
Outlet_Identifier               0
Outlet_Establishment_Year       0
Outlet_Size                  2410
Outlet_Location_Type            0
Outlet_Type                     0
Item_Outlet_Sales               0
dtype: int64


Salestest=pd.read_csv("Test.csv")


Salestest


#### Checking the dimension of the dataset
Salestest.shape

(5681, 11)


Salestest.head(10)


Salestest.tail(10)


ProfileReport(Salestest)

Summarize dataset:   0%|          | 0/5 [00:00<?, ?it/s]

Generate report structure:   0%|          | 0/1 [00:00<?, ?it/s]

Render HTML:   0%|          | 0/1 [00:00<?, ?it/s]


Salestest.describe()


Salestest.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 5681 entries, 0 to 5680
Data columns (total 11 columns):
 #   Column                     Non-Null Count  Dtype  
---  ------                     --------------  -----  
 0   Item_Identifier            5681 non-null   object 
 1   Item_Weight                4705 non-null   float64
 2   Item_Fat_Content           5681 non-null   object 
 3   Item_Visibility            5681 non-null   float64
 4   Item_Type                  5681 non-null   object 
 5   Item_MRP                   5681 non-null   float64
 6   Outlet_Identifier          5681 non-null   object 
 7   Outlet_Establishment_Year  5681 non-null   int64  
 8   Outlet_Size                4075 non-null   object 
 9   Outlet_Location_Type       5681 non-null   object 
 10  Outlet_Type                5681 non-null   object 
dtypes: float64(3), int64(1), object(7)
memory usage: 488.3+ KB


Salestest.isna().sum()

Item_Identifier                 0
Item_Weight                   976
Item_Fat_Content                0
Item_Visibility                 0
Item_Type                       0
Item_MRP                        0
Outlet_Identifier               0
Outlet_Establishment_Year       0
Outlet_Size                  1606
Outlet_Location_Type            0
Outlet_Type                     0
dtype: int64


x=Salestrain.drop('Item_Weight',axis=1)
for i in x.columns:
    H=x[i]
    H=np.array(H)
    Counter(H)
    J=[]
    entropy=0
    for j in Counter(H):
        p=Counter(H)[j]
        J==J.append(p)
    for k in J:
        a_k=-(k/len(H))*(math.log(k/len(H),len(J)))
        entropy=entropy+a_k
    print (f"The entropy for {i} is {entropy}")

The entropy for Item_Identifier is 0.9943672571855067
The entropy for Item_Fat_Content is 0.5670159638726668
The entropy for Item_Visibility is 0.9635001925885082
The entropy for Item_Type is 0.912856148641514
The entropy for Item_MRP is 0.9879625726182639
The entropy for Outlet_Identifier is 0.9919373055756733
The entropy for Outlet_Establishment_Year is 0.9884150951444971
The entropy for Outlet_Size is 0.9530930510010018
The entropy for Outlet_Location_Type is 0.991258113296581
The entropy for Outlet_Type is 0.7383363040319846
The entropy for Item_Outlet_Sales is 0.9687158802197017


sns.kdeplot(x=Salestrain['Item_Weight'])

<AxesSubplot:xlabel='Item_Weight', ylabel='Density'>


x=Salestest.drop('Item_Weight',axis=1)
for i in x.columns:
    H=x[i]
    H=np.array(H)
    Counter(H)
    J=[]
    entropy=0
    for j in Counter(H):
        p=Counter(H)[j]
        J==J.append(p)
    for k in J:
        a_k=-(k/len(H))*(math.log(k/len(H),len(J)))
        entropy=entropy+a_k
    print (f"The entropy for {i} is {entropy}")

The entropy for Item_Identifier is 0.9890062352875344
The entropy for Item_Fat_Content is 0.5625153732709652
The entropy for Item_Visibility is 0.9645968858582553
The entropy for Item_Type is 0.9137299613085187
The entropy for Item_MRP is 0.9903016313311045
The entropy for Outlet_Identifier is 0.9919420219573883
The entropy for Outlet_Establishment_Year is 0.9883871445154377
The entropy for Outlet_Size is 0.9530602028397074
The entropy for Outlet_Location_Type is 0.9912632408343459
The entropy for Outlet_Type is 0.7384116845306994


sns.kdeplot(x=Salestest['Item_Weight'])

<AxesSubplot:xlabel='Item_Weight', ylabel='Density'>


plt.figure(figsize=(18,10))

plt.subplot(2,2,1)
plt.title('Item_Weight in Train dataset')
sns.boxplot(x='Item_Weight', data=Salestrain)

plt.subplot(2,2,2)
plt.title('Item_Weight in Test dataset')
sns.boxplot(x='Item_Weight', data=Salestest)

plt.show()


# Use the mean to replace the null values in Item_Weight feature
Salestrain.Item_Weight = Salestrain.Item_Weight.fillna(Salestrain.Item_Weight.mean())
Salestest.Item_Weight = Salestest.Item_Weight.fillna(Salestest.Item_Weight.mean())


Salestrain.mode()


Salestrain.Outlet_Size.unique()

array(['Medium', nan, 'High', 'Small'], dtype=object)


Salestrain.Outlet_Size.value_counts()

Medium    2793
Small     2388
High       932
Name: Outlet_Size, dtype: int64


Salestest.Outlet_Size.value_counts()

Medium    1862
Small     1592
High       621
Name: Outlet_Size, dtype: int64


Salestrain.Outlet_Size = Salestrain.Outlet_Size.fillna(Salestrain.Outlet_Size.mode()[0])
Salestest.Outlet_Size = Salestest.Outlet_Size.fillna(Salestest.Outlet_Size.mode()[0])


Salestrain.isnull().sum()

Item_Identifier              0
Item_Weight                  0
Item_Fat_Content             0
Item_Visibility              0
Item_Type                    0
Item_MRP                     0
Outlet_Identifier            0
Outlet_Establishment_Year    0
Outlet_Size                  0
Outlet_Location_Type         0
Outlet_Type                  0
Item_Outlet_Sales            0
dtype: int64


Salestest.isnull().sum()

Item_Identifier              0
Item_Weight                  0
Item_Fat_Content             0
Item_Visibility              0
Item_Type                    0
Item_MRP                     0
Outlet_Identifier            0
Outlet_Establishment_Year    0
Outlet_Size                  0
Outlet_Location_Type         0
Outlet_Type                  0
dtype: int64


Salestrain.nunique()

Item_Identifier              1559
Item_Weight                   416
Item_Fat_Content                5
Item_Visibility              7880
Item_Type                      16
Item_MRP                     5938
Outlet_Identifier              10
Outlet_Establishment_Year       9
Outlet_Size                     3
Outlet_Location_Type            3
Outlet_Type                     4
Item_Outlet_Sales            3493
dtype: int64


Salestrain.Item_Fat_Content.unique()

array(['Low Fat', 'Regular', 'low fat', 'LF', 'reg'], dtype=object)


Salestrain.Item_Fat_Content.value_counts()

Low Fat    5089
Regular    2889
LF          316
reg         117
low fat     112
Name: Item_Fat_Content, dtype: int64


Salestrain.Item_Fat_Content = Salestrain.Item_Fat_Content.replace(to_replace=['low fat', 'LF', 'reg'], value=['Low Fat', 'Low Fat', 'Regular'])


Salestrain.Item_Fat_Content.value_counts()

Low Fat    5517
Regular    3006
Name: Item_Fat_Content, dtype: int64


Salestest.Item_Fat_Content.unique()

array(['Low Fat', 'reg', 'Regular', 'LF', 'low fat'], dtype=object)


Salestest.Item_Fat_Content.value_counts()

Low Fat    3396
Regular    1935
LF          206
reg          78
low fat      66
Name: Item_Fat_Content, dtype: int64


Salestest.Item_Fat_Content = Salestest.Item_Fat_Content.replace(to_replace=['low fat', 'LF', 'reg'], value=['Low Fat', 'Low Fat', 'Regular'])


Salestest.Item_Fat_Content.value_counts()

Low Fat    3668
Regular    2013
Name: Item_Fat_Content, dtype: int64


Salestest.Item_Fat_Content.unique()

array(['Low Fat', 'Regular'], dtype=object)


Salestest.Item_Fat_Content.value_counts()

Low Fat    3668
Regular    2013
Name: Item_Fat_Content, dtype: int64


plt.figure(figsize=(18,12))

plt.subplot(2,2,1)
plt.title("Item_Fat_Content Count in the Training Data")
sns.countplot(x='Item_Fat_Content', data=Salestrain, palette='ocean')

plt.subplot(2,2,2)
low_fat = Salestrain.Item_Fat_Content[Salestrain.Item_Fat_Content == 'Low Fat'].count()
reg = Salestrain.Item_Fat_Content[Salestrain.Item_Fat_Content == 'Regular'].count()
plt.title("Item Content Fat Distrbution in the Training Data")
plt.pie([low_fat, reg], labels=['Low Fat', 'Regular'], explode=[0.01,0.01], autopct="%.2f%%", colors=['lightgreen', 'yellow'], shadow=True)
plt.legend()

plt.show()


plt.figure(figsize=(25,8))
plt.title("Item Type count in the Training data", {"fontsize" : 25})
order = Salestrain.groupby('Item_Type').count().sort_values(by='Item_Outlet_Sales', ascending=False).index
sns.countplot(x='Item_Type', data=Salestrain, order=order);


Salestrain.Outlet_Size.value_counts()

Medium    5203
Small     2388
High       932
Name: Outlet_Size, dtype: int64


plt.figure(figsize=(18,12))

plt.subplot(2,2,1)
plt.title("The size of the store in terms of ground area covered count")
sns.countplot(x='Outlet_Size', data=Salestrain, palette='spring')

plt.subplot(2,2,2)
medium = Salestrain.Outlet_Size[Salestrain.Outlet_Size == 'Medium'].count()
small = Salestrain.Outlet_Size[Salestrain.Outlet_Size == 'Small'].count()
high = Salestrain.Outlet_Size[Salestrain.Outlet_Size == 'High'].count()
plt.title("The size of the store in terms of ground area covered Distrbution")
plt.pie([medium, small, high], labels=['Medium', 'Small', 'High'], autopct="%.2f%%", shadow=True, explode=[0.01,0.01,0.01])
my_circle = plt.Circle( (0,0), 0.4, color='white')
plt.gcf().gca().add_artist(my_circle)
plt.legend()

plt.show()


Salestrain.Outlet_Location_Type.unique()

array(['Tier 1', 'Tier 3', 'Tier 2'], dtype=object)


Salestrain.Outlet_Location_Type.value_counts()

Tier 3    3350
Tier 2    2785
Tier 1    2388
Name: Outlet_Location_Type, dtype: int64


plt.figure(figsize=(18,12))

plt.subplot(2,2,1)
plt.title("The type of city in which the store is located")
sns.countplot(x='Outlet_Location_Type', data=Salestrain, palette='twilight')

plt.subplot(2,2,2)
tier_1 = Salestrain.Outlet_Location_Type[Salestrain.Outlet_Location_Type == 'Tier 1'].count()
tier_2 = Salestrain.Outlet_Location_Type[Salestrain.Outlet_Location_Type == 'Tier 2'].count()
tier_3 = Salestrain.Outlet_Location_Type[Salestrain.Outlet_Location_Type == 'Tier 3'].count()
plt.title("The type of city in which the store is located Distrbution")
plt.pie([tier_1, tier_2, tier_3], labels=['tier_1', 'tier_2', 'tier_3'], autopct="%.2f%%", shadow=True, explode=[0.05,0.05,0.05])
my_circle = plt.Circle( (0,0), 0.4, color='white')
plt.gcf().gca().add_artist(my_circle)
plt.legend()

plt.show()


Salestrain.Outlet_Type.unique()

array(['Supermarket Type1', 'Supermarket Type2', 'Grocery Store',
       'Supermarket Type3'], dtype=object)


Salestrain.Outlet_Type.value_counts()

Supermarket Type1    5577
Grocery Store        1083
Supermarket Type3     935
Supermarket Type2     928
Name: Outlet_Type, dtype: int64


plt.figure(figsize=(18,12))

plt.subplot(2,2,1)
plt.title("Whether the outlet is just a grocery store or some sort of supermarket count")
sns.countplot(x='Outlet_Type', data=Salestrain, palette='autumn')

plt.subplot(2,2,2)
sup_1 = Salestrain.Outlet_Type[Salestrain.Outlet_Type == 'Supermarket Type1'].count()
goc = Salestrain.Outlet_Type[Salestrain.Outlet_Type == 'Grocery Store'].count()
sup_3 = Salestrain.Outlet_Type[Salestrain.Outlet_Type == 'Supermarket Type3'].count()
sup_2 = Salestrain.Outlet_Type[Salestrain.Outlet_Type == 'Supermarket Type2'].count()
plt.title("Whether the outlet is just a grocery store or some sort of supermarket Distrbution")
plt.pie([sup_1, goc, sup_3, sup_2], labels=['Supermarket Type1', 'Grocery Store', 'Supermarket Type3', 'Supermarket Type2'], autopct="%.2f%%", shadow=True, explode=[0.01,0.01,0.01, 0.01])
my_circle = plt.Circle( (0,0), 0.4, color='white')
plt.gcf().gca().add_artist(my_circle)
plt.legend(bbox_to_anchor=(1, 0))

plt.show()


Salestrain.Outlet_Establishment_Year.unique()

array([1999, 2009, 1998, 1987, 1985, 2002, 2007, 1997, 2004], dtype=int64)


Salestrain['Age'] = 2023 - Salestrain.Outlet_Establishment_Year
Salestest['Age'] = 2023 - Salestest.Outlet_Establishment_Year


plt.figure(figsize=(12,6))

sns.countplot(x='Age', data=Salestrain, palette='autumn');


Salestrain.nunique()

Item_Identifier              1559
Item_Weight                   416
Item_Fat_Content                2
Item_Visibility              7880
Item_Type                      16
Item_MRP                     5938
Outlet_Identifier              10
Outlet_Establishment_Year       9
Outlet_Size                     3
Outlet_Location_Type            3
Outlet_Type                     4
Item_Outlet_Sales            3493
Age                             9
dtype: int64


FEATURES = [col for col in Salestrain.columns if col not in ['Item_Outlet_Sales', 'Outlet_Identifier']]
cat_features = [col for col in FEATURES if Salestrain[col].nunique() < 10]
cont_features = [col for col in FEATURES if Salestrain[col].nunique() >= 10]


cat_features

['Item_Fat_Content',
 'Outlet_Establishment_Year',
 'Outlet_Size',
 'Outlet_Location_Type',
 'Outlet_Type',
 'Age']


cont_features

['Item_Identifier', 'Item_Weight', 'Item_Visibility', 'Item_Type', 'Item_MRP']


ncols = 3
nrows = int(len(cat_features) / ncols + (len(FEATURES) % ncols > 0))-1

fig, axes = plt.subplots(nrows, ncols, figsize=(22, 10), facecolor='#EAEAF2')

for r in range(nrows):
    for c in range(ncols):
        col = cat_features[r*ncols+c]
        sns.barplot(y=Salestrain['Item_Outlet_Sales'], x=Salestrain[col], ax=axes[r, c], palette='autumn', label='Train data')
        axes[r, c].set_ylabel('')
        axes[r, c].set_xlabel(col, fontsize=12, fontweight='bold')
        axes[r, c].tick_params(labelsize=10, width=0.5)
        axes[r, c].xaxis.offsetText.set_fontsize(6)
        axes[r, c].yaxis.offsetText.set_fontsize(6)
plt.show()


plt.figure(figsize=(25,8))

order=Salestrain.groupby('Item_Type').mean().sort_values(by='Item_Outlet_Sales', ascending=False).index
sns.barplot(x='Item_Type',y='Item_Outlet_Sales',data=Salestrain,palette='spring', order=order);


Salestrain.nunique()

Item_Identifier              1559
Item_Weight                   416
Item_Fat_Content                2
Item_Visibility              7880
Item_Type                      16
Item_MRP                     5938
Outlet_Identifier              10
Outlet_Establishment_Year       9
Outlet_Size                     3
Outlet_Location_Type            3
Outlet_Type                     4
Item_Outlet_Sales            3493
Age                             9
dtype: int64


FEATURES = [col for col in Salestrain.columns if col not in ['Item_Outlet_Sales', 'Outlet_Identifier', 'Item_Identifier']]
cat_features = [col for col in FEATURES if Salestrain[col].nunique() < 25]
cont_features = [col for col in FEATURES if Salestrain[col].nunique() >= 25]


cont_features

['Item_Weight', 'Item_Visibility', 'Item_MRP']


# distrbution of the continous features
plt.figure(figsize=(10,20))
plt.subplot(4,1,1)
plt.title("Item_Weight Distrbution")
sns.kdeplot(x='Item_Weight', data=Salestrain)

plt.subplot(4,1,2)
plt.title("Item_Visibility Distrbution")
sns.kdeplot(x='Item_Visibility', data=Salestrain);

plt.subplot(4,1,3)
plt.title("Item_MRP Distrbution")
sns.kdeplot(x='Item_MRP', data=Salestrain);

plt.subplot(4,1,4)
plt.title("Item_Outlet_Sales")
sns.kdeplot(x='Item_Outlet_Sales', data=Salestrain);


plt.figure(figsize=(18,15))
plt.subplot(3,1,1)
sns.lineplot(x='Item_Weight', y='Item_Outlet_Sales', data=Salestrain);

plt.subplot(3,1,2)
sns.lineplot(x='Item_Visibility', y='Item_Outlet_Sales', data=Salestrain);

plt.subplot(3,1,3)
sns.lineplot(x='Item_MRP', y='Item_Outlet_Sales', data=Salestrain);


plt.figure(figsize=(25,5))
sns.barplot('Item_Type','Item_Outlet_Sales',hue='Item_Fat_Content',data=Salestrain,palette='mako')
plt.legend();


plt.figure(figsize=(10,5))
sns.barplot('Outlet_Location_Type','Item_Outlet_Sales',hue='Outlet_Type',data=Salestrain,palette='magma')
plt.legend()
plt.show()


Salestrain.nunique()

Item_Identifier              1559
Item_Weight                   416
Item_Fat_Content                2
Item_Visibility              7880
Item_Type                      16
Item_MRP                     5938
Outlet_Identifier              10
Outlet_Establishment_Year       9
Outlet_Size                     3
Outlet_Location_Type            3
Outlet_Type                     4
Item_Outlet_Sales            3493
Age                             9
dtype: int64


plt.figure(figsize=(10,5))
sns.barplot('Item_Fat_Content','Item_Outlet_Sales',hue='Outlet_Size',data=Salestrain,palette='autumn')
plt.legend()
plt.show()


plt.figure(figsize=(10,5))
sns.barplot('Item_Fat_Content','Item_Outlet_Sales',hue='Outlet_Type',data=Salestrain,palette='summer')
plt.legend()
plt.show()


missing_cols = []
for column in Salestrain.loc[:, ~Salestrain.columns.isin(['Item_Outlet_Sales'])]:
    if column not in Salestest.columns:
        missing_cols.append()
print("Number of features present in training data set that are not in test data:", len(missing_cols))      

missing_cols = []
for column in Salestest.loc[:,~Salestest.columns.isin(['Item_Identifier'])]:
    if column not in Salestrain.columns:
        missing_cols.append()
print("Number of features present in test data set that are not in training data:", len(missing_cols))

Number of features present in training data set that are not in test data: 0
Number of features present in test data set that are not in training data: 0


Salestrain.nunique()

Item_Identifier              1559
Item_Weight                   416
Item_Fat_Content                2
Item_Visibility              7880
Item_Type                      16
Item_MRP                     5938
Outlet_Identifier              10
Outlet_Establishment_Year       9
Outlet_Size                     3
Outlet_Location_Type            3
Outlet_Type                     4
Item_Outlet_Sales            3493
Age                             9
dtype: int64


Salestrain.dtypes

Item_Identifier               object
Item_Weight                  float64
Item_Fat_Content              object
Item_Visibility              float64
Item_Type                     object
Item_MRP                     float64
Outlet_Identifier             object
Outlet_Establishment_Year      int64
Outlet_Size                   object
Outlet_Location_Type          object
Outlet_Type                   object
Item_Outlet_Sales            float64
Age                            int64
dtype: object


Salestest.dtypes

Item_Identifier               object
Item_Weight                  float64
Item_Fat_Content              object
Item_Visibility              float64
Item_Type                     object
Item_MRP                     float64
Outlet_Identifier             object
Outlet_Establishment_Year      int64
Outlet_Size                   object
Outlet_Location_Type          object
Outlet_Type                   object
Age                            int64
dtype: object


Salestrain.columns[Salestrain.isna().any()]

Index([], dtype='object')


Salestest


Salestrain.dtypes

Item_Identifier               object
Item_Weight                  float64
Item_Fat_Content              object
Item_Visibility              float64
Item_Type                     object
Item_MRP                     float64
Outlet_Identifier             object
Outlet_Establishment_Year      int64
Outlet_Size                   object
Outlet_Location_Type          object
Outlet_Type                   object
Item_Outlet_Sales            float64
Age                            int64
dtype: object


Salestest.dtypes

Item_Identifier               object
Item_Weight                  float64
Item_Fat_Content              object
Item_Visibility              float64
Item_Type                     object
Item_MRP                     float64
Outlet_Identifier             object
Outlet_Establishment_Year      int64
Outlet_Size                   object
Outlet_Location_Type          object
Outlet_Type                   object
Age                            int64
dtype: object


#drop non needed features
Salestrain = Salestrain.drop(['Item_Identifier','Outlet_Identifier','Outlet_Establishment_Year'],axis=1)
Salestest= Salestest.drop(['Item_Identifier','Outlet_Identifier','Outlet_Establishment_Year'],axis=1)


Salestrain.corr()


plt.figure(figsize=(20,10))
sns.heatmap(Salestrain.corr(), annot=True);


Salestrain.head()


Salestest.head()


cont_features

['Item_Weight', 'Item_Visibility', 'Item_MRP']


Salestrain.head()


Salestest.head()


X = Salestrain[cont_features]
X.head()


Salestest2 = Salestest[cont_features]
Salestest2.head()


# Calculating VIF for all numerical variables
from statsmodels.stats.outliers_influence import variance_inflation_factor

x = Salestrain[['Item_Weight', 'Item_Visibility', 'Item_MRP']]

vif = pd.DataFrame()
vif["features"] = x.columns
vif["vif_Factor"] = [variance_inflation_factor(x.values, i) for i in range(x.shape[1])]
vif


Salestrain.skew()

Item_Weight          0.090561
Item_Visibility      1.167091
Item_MRP             0.127202
Item_Outlet_Sales    1.177531
Age                  0.396641
dtype: float64


# Before Scaling
fig, ax = plt.subplots(2,3,figsize = (9,9))
sns.distplot(Salestrain["Item_Visibility"], kde =True, ax=ax[0,0], color = "red")
sns.distplot(Salestrain["Item_MRP"], kde =True, ax=ax[0,1], color = "blue")
sns.distplot(Salestrain["Age"], kde =True, ax=ax[0,2], color = "orange")
sns.distplot(Salestrain["Item_Outlet_Sales"], kde =True, ax=ax[1,0], color = "magenta")
sns.distplot(Salestrain["Item_Weight"], kde =True, ax=ax[1,1], color = "black")

<AxesSubplot:xlabel='Item_Weight', ylabel='Density'>


cat_train = Salestrain.select_dtypes(include=['object']).columns
cat_test = Salestest.select_dtypes(include=['object']).columns

num_train = Salestrain.select_dtypes(exclude = ['object']).columns
num_test = Salestest.select_dtypes(exclude = ['object']).columns

print("Numerical features : " + str(len(num_train)))
print("Categorical features : " + str(len(cat_train)))

feat_cat_train = Salestrain[cat_train]
feat_cat_test = Salestest[cat_test]
feat_num_train = Salestrain[num_train]
feat_num_test = Salestest[num_test]

Numerical features : 5
Categorical features : 5


from scipy.stats import skew 
skewness = feat_num_train.apply(lambda x: skew(x))
skewness.sort_values(ascending=False)

Item_Outlet_Sales    1.177323
Item_Visibility      1.166885
Age                  0.396571
Item_MRP             0.127180
Item_Weight          0.090546
dtype: float64


## Transforming features in train data 

skewness = skewness[abs(skewness) > 0.50]
skewed_index = skewness.index

for i in skewed_index:
    feat_num_train[i] = np.log(feat_num_train[i]+1)


## Transforming features in test data for the same features

for i in skewed_index[skewed_index != 'Item_Outlet_Sales']:
    feat_num_test[i] = np.log(feat_num_test[i]+1)


y = feat_num_train['Item_Outlet_Sales']
feat_num_train = feat_num_train.drop("Item_Outlet_Sales",axis=1)


from sklearn.preprocessing import OneHotEncoder

OH_encoder = OneHotEncoder(handle_unknown='ignore', sparse = False)
OH_train = pd.DataFrame(OH_encoder.fit_transform(feat_cat_train))
OH_test = pd.DataFrame(OH_encoder.transform(feat_cat_test))

# One-hot encoding removed index; put it back
OH_train.index = feat_cat_train.index
OH_test.index = feat_cat_test.index


from sklearn.feature_selection import RFE
from sklearn.tree import DecisionTreeRegressor

rfe = RFE(estimator = DecisionTreeRegressor(),n_features_to_select=20)
rfe.fit(feat_num_train, y)

selected_feat_RFE = feat_num_train.columns[(rfe.get_support())]

# let's print some stats
print('Total features: {}'.format((feat_num_train.shape[1])))
print('Selected features: {}'.format(len(selected_feat_RFE)))
print('The selected features are: {}'.format(selected_feat_RFE))

Total features: 4
Selected features: 4
The selected features are: Index(['Item_Weight', 'Item_Visibility', 'Item_MRP', 'Age'], dtype='object')


X = pd.concat([OH_train,feat_num_train],axis = 1)
X_test = pd.concat([OH_test,feat_num_test], axis = 1)
X.columns = X.columns.astype(str)
X_test.columns = X_test.columns.astype(str)


x_train, x_test, y_train, y_test = train_test_split(X, y, test_size=0.20, random_state=0)


import statsmodels.api as sm
# Fit the linear regression model
model = sm.OLS(y_train, x_train).fit()


y_pred=model.predict(x_test)


rmse_lin = np.sqrt(metrics.mean_squared_error(y_test, y_pred))
r2_score = metrics.r2_score(y_test, y_pred)
print('Root Mean Squared Error:',rmse_lin )
print('R2_Score:',r2_score*100,"%" )

Root Mean Squared Error: 0.5515520555642495
R2_Score: 71.42586803830434 %


residue_lreg = y_test - y_pred
#Plotting Residual Plot
plt.scatter(y_test,residue_lreg, c = "brown")
plt.xlabel("Residual Plot for Linear Regression")
plt.ylabel("y_test")
plt.axhline(y = 0)

<matplotlib.lines.Line2D at 0x2356e787850>


# Hyper-paramaters to be used during tuning. The specific range for each hyper-parameter 
# was selected after carrying out several iterations.
e_alphas = [0.0001, 0.0002, 0.0003, 0.0004, 0.0005, 0.0006, 0.0007]
e_l1ratio = [0.8, 0.85, 0.9, 0.95, 0.99, 1]
alphas_ridge = [14.5, 14.6, 14.7, 14.8, 14.9, 15, 15.1, 15.2, 15.3, 15.4, 15.5]
alphas_lasso = [0.0001, 0.0002, 0.0003, 0.0004, 0.0005, 0.0006, 0.0007, 0.0008]


#Create model scoring and kfolds functions

kfolds = KFold(n_splits=10)


from sklearn.linear_model import ElasticNet

# Create an instance of the ElasticNet model
elastic_net = ElasticNetCV(alphas=e_alphas, cv=kfolds, l1_ratio=e_l1ratio)

# Fit the model to your training data
elastic_net.fit(x_train, y_train)

# Predictions
predictions = elastic_net.predict(x_test)


rmse_EN = np.sqrt(metrics.mean_squared_error(y_test, predictions))
en_score = metrics.r2_score(y_test, predictions)
print('Root Mean Squared Error:',rmse_EN )
print('R2_Score:',en_score*100,"%" )

Root Mean Squared Error: 0.5513000874440245
R2_Score: 71.45196938519395 %


residue_lreg = y_test - predictions
#Plotting Residual Plot
plt.scatter(y_test,residue_lreg, c = "brown")
plt.xlabel("Residual Plot for Elastic Net Regression")
plt.ylabel("y_test")
plt.axhline(y = 0)

<matplotlib.lines.Line2D at 0x2356fe2e7c0>


#fitting the model
from sklearn.linear_model import Lasso, Ridge
ls = LassoCV(alphas=alphas_lasso, cv=kfolds)
ls = ls.fit(x_train, y_train)


ls_pred=ls.predict(x_test)


from sklearn.metrics import r2_score
rmse = np.sqrt(metrics.mean_squared_error(y_test, ls_pred))
r2_score_LS = r2_score(y_test, ls_pred)
print('R2_Score:',r2_score_LS*100,"%" )
print('Root Mean Squared Error:',rmse )

R2_Score: 71.45588830765148 %
Root Mean Squared Error: 0.5512622463671155


residue_lreg = y_test - y_pred
#Plotting Residual Plot
plt.scatter(y_test,residue_lreg, c = "brown")
plt.xlabel("Residual Plot for Lasso Regression")
plt.ylabel("y_test")
plt.axhline(y = 0)

<matplotlib.lines.Line2D at 0x2356fe2ed30>


#Ridge Regression
rr = RidgeCV(alphas=alphas_ridge, cv=kfolds)
rr.fit(x_train, y_train)

RidgeCV(alphas=[14.5, 14.6, 14.7, 14.8, 14.9, 15, 15.1, 15.2, 15.3, 15.4, 15.5],
        cv=KFold(n_splits=10, random_state=None, shuffle=False))

RidgeCV(alphas=[14.5, 14.6, 14.7, 14.8, 14.9, 15, 15.1, 15.2, 15.3, 15.4, 15.5],
        cv=KFold(n_splits=10, random_state=None, shuffle=False))


#Prediction AFTER Ridge regression
rr_pred = rr.predict(x_test)


#Accuracy score check and RMSE
rmse_ridge = np.sqrt(metrics.mean_squared_error(y_test, rr_pred))
print('Root Mean Squared Error:',rmse_ridge )
r2_score_RR = r2_score(y_test, y_pred)
print('R2_Score:',r2_score_RR*100,"%" )

Root Mean Squared Error: 0.5517871779846749
R2_Score: 71.42586803830434 %


#residue after ridge
residue_rr = y_test-rr_pred
#Plotting Residual Plot
plt.scatter(y_test,residue_rr, c = "blue")
plt.xlabel("Residual Plot for Ridge Regression")
plt.ylabel("y_test")
plt.axhline(y = 0)

<matplotlib.lines.Line2D at 0x2356e229af0>


# Create a Decision Tree regressor model
dt_model = DecisionTreeRegressor(random_state=42)

# Train the model on the training data
dt_model.fit(x_train, y_train)

# Make predictions on the test data
dt_pred = dt_model.predict(x_test)


# Evaluate the model
rmse_dt = np.sqrt(metrics.mean_squared_error(y_test, dt_pred))
print('Root Mean Squared Error:',rmse_dt )
r2_dt = metrics.r2_score(y_test, dt_pred)
print('R-squared:', r2_dt)

Root Mean Squared Error: 0.7577915929905223
R-squared: 0.4606140908409836


#residue after decision tree
residue_dt = y_test-dt_pred
#Plotting Residual Plot
plt.scatter(y_test,residue_dt, c = "yellow")
plt.xlabel("Residual Plot for Decision Tree Regression")
plt.ylabel("y_test")
plt.axhline(y = 0)

<matplotlib.lines.Line2D at 0x2357185d520>


# Create a Random Forest regressor model
rf_model = RandomForestRegressor(n_estimators=100, random_state=42)

# Train the model on the training data
rf_model.fit(x_train, y_train)

# Make predictions on the test data
rf_pred = rf_model.predict(x_test)

# Evaluate the model
rmse_rf = np.sqrt(metrics.mean_squared_error(y_test, rf_pred))
print('Root Mean Squared Error:',rmse_rf )
r2_rf = metrics.r2_score(y_test, rf_pred)
print('R-squared:', r2_rf)

#residue after random forest
residue_rf = y_test-rf_pred

#Plotting Residual Plot
plt.scatter(y_test,residue_rf, c = "green")
plt.xlabel("Residual Plot for Random Forest Regression")
plt.ylabel("y_test")
plt.axhline(y = 0)

Root Mean Squared Error: 0.5678768756082895
R-squared: 0.6970936348217647

<matplotlib.lines.Line2D at 0x2356ffce3d0>


pd.DataFrame(data = {'Random Forest': r2_rf, "Lasso Regression": r2_score_LS,"Ridge Regression":r2_score_RR,'Linear Regression': r2_score, 'Decision Tree':r2_dt, 'Elastic Net': en_score},index = ['R2_Score'])


scoring = pd.DataFrame({'Ridge': rmse_ridge,"Lasso": rmse,"ElasticNet": rmse_EN,"LinReg" : rmse_lin,"RandomForest" : rmse_rf,"DecisionTree": rmse_dt}, index = ['RMSE'])
scoring


plt.figure(figsize=(8,6))
plt.title('Root Mean Square plot of different models')
sns.barplot(data = scoring)
plt.xlabel("Model")
plt.ylabel("RMSE Score")

Text(0, 0.5, 'RMSE Score')


SalesPredicted=elastic_net.predict(X_test)


#Performing inverse transformation
actual_item_outsale = np.exp(SalesPredicted)-1


actual_item_outsale

array([1373.36123556, 1120.38912505,  601.48476623, ..., 1547.18952135,
       3552.7666112 , 1155.9846947 ])


ItemSales=pd.DataFrame(actual_item_outsale)


Salestest=Salestest.join(ItemSales)


Salestest=Salestest.rename(columns={0:"Predicted Sales"})


Salestest.to_csv('SalesPred.csv')

	Item_Identifier	Item_Weight	Item_Fat_Content	Item_Visibility	Item_Type	Item_MRP	Outlet_Identifier	Outlet_Establishment_Year	Outlet_Size	Outlet_Location_Type	Outlet_Type	Item_Outlet_Sales
0	FDA15	9.300	Low Fat	0.016047	Dairy	249.8092	OUT049	1999	Medium	Tier 1	Supermarket Type1	3735.1380
1	DRC01	5.920	Regular	0.019278	Soft Drinks	48.2692	OUT018	2009	Medium	Tier 3	Supermarket Type2	443.4228
2	FDN15	17.500	Low Fat	0.016760	Meat	141.6180	OUT049	1999	Medium	Tier 1	Supermarket Type1	2097.2700
3	FDX07	19.200	Regular	0.000000	Fruits and Vegetables	182.0950	OUT010	1998	NaN	Tier 3	Grocery Store	732.3800
4	NCD19	8.930	Low Fat	0.000000	Household	53.8614	OUT013	1987	High	Tier 3	Supermarket Type1	994.7052
5	FDP36	10.395	Regular	0.000000	Baking Goods	51.4008	OUT018	2009	Medium	Tier 3	Supermarket Type2	556.6088
6	FDO10	13.650	Regular	0.012741	Snack Foods	57.6588	OUT013	1987	High	Tier 3	Supermarket Type1	343.5528
7	FDP10	NaN	Low Fat	0.127470	Snack Foods	107.7622	OUT027	1985	Medium	Tier 3	Supermarket Type3	4022.7636
8	FDH17	16.200	Regular	0.016687	Frozen Foods	96.9726	OUT045	2002	NaN	Tier 2	Supermarket Type1	1076.5986
9	FDU28	19.200	Regular	0.094450	Frozen Foods	187.8214	OUT017	2007	NaN	Tier 2	Supermarket Type1	4710.5350

	Item_Identifier	Item_Weight	Item_Fat_Content	Item_Visibility	Item_Type	Item_MRP	Outlet_Identifier	Outlet_Establishment_Year	Outlet_Size	Outlet_Location_Type	Outlet_Type	Item_Outlet_Sales
8513	FDH31	12.000	Regular	0.020407	Meat	99.9042	OUT035	2004	Small	Tier 2	Supermarket Type1	595.2252
8514	FDA01	15.000	Regular	0.054489	Canned	57.5904	OUT045	2002	NaN	Tier 2	Supermarket Type1	468.7232
8515	FDH24	20.700	Low Fat	0.021518	Baking Goods	157.5288	OUT018	2009	Medium	Tier 3	Supermarket Type2	1571.2880
8516	NCJ19	18.600	Low Fat	0.118661	Others	58.7588	OUT018	2009	Medium	Tier 3	Supermarket Type2	858.8820
8517	FDF53	20.750	reg	0.083607	Frozen Foods	178.8318	OUT046	1997	Small	Tier 1	Supermarket Type1	3608.6360
8518	FDF22	6.865	Low Fat	0.056783	Snack Foods	214.5218	OUT013	1987	High	Tier 3	Supermarket Type1	2778.3834
8519	FDS36	8.380	Regular	0.046982	Baking Goods	108.1570	OUT045	2002	NaN	Tier 2	Supermarket Type1	549.2850
8520	NCJ29	10.600	Low Fat	0.035186	Health and Hygiene	85.1224	OUT035	2004	Small	Tier 2	Supermarket Type1	1193.1136
8521	FDN46	7.210	Regular	0.145221	Snack Foods	103.1332	OUT018	2009	Medium	Tier 3	Supermarket Type2	1845.5976
8522	DRG01	14.800	Low Fat	0.044878	Soft Drinks	75.4670	OUT046	1997	Small	Tier 1	Supermarket Type1	765.6700

	Item_Weight	Item_Visibility	Item_MRP	Outlet_Establishment_Year	Item_Outlet_Sales
count	7060.000000	8523.000000	8523.000000	8523.000000	8523.000000
mean	12.857645	0.066132	140.992782	1997.831867	2181.288914
std	4.643456	0.051598	62.275067	8.371760	1706.499616
min	4.555000	0.000000	31.290000	1985.000000	33.290000
25%	8.773750	0.026989	93.826500	1987.000000	834.247400
50%	12.600000	0.053931	143.012800	1999.000000	1794.331000
75%	16.850000	0.094585	185.643700	2004.000000	3101.296400
max	21.350000	0.328391	266.888400	2009.000000	13086.964800

	Item_Identifier	Item_Weight	Item_Fat_Content	Item_Visibility	Item_Type	Item_MRP	Outlet_Identifier	Outlet_Establishment_Year	Outlet_Size	Outlet_Location_Type	Outlet_Type
0	FDW58	20.750	Low Fat	0.007565	Snack Foods	107.8622	OUT049	1999	Medium	Tier 1	Supermarket Type1
1	FDW14	8.300	reg	0.038428	Dairy	87.3198	OUT017	2007	NaN	Tier 2	Supermarket Type1
2	NCN55	14.600	Low Fat	0.099575	Others	241.7538	OUT010	1998	NaN	Tier 3	Grocery Store
3	FDQ58	7.315	Low Fat	0.015388	Snack Foods	155.0340	OUT017	2007	NaN	Tier 2	Supermarket Type1
4	FDY38	NaN	Regular	0.118599	Dairy	234.2300	OUT027	1985	Medium	Tier 3	Supermarket Type3
5	FDH56	9.800	Regular	0.063817	Fruits and Vegetables	117.1492	OUT046	1997	Small	Tier 1	Supermarket Type1
6	FDL48	19.350	Regular	0.082602	Baking Goods	50.1034	OUT018	2009	Medium	Tier 3	Supermarket Type2
7	FDC48	NaN	Low Fat	0.015782	Baking Goods	81.0592	OUT027	1985	Medium	Tier 3	Supermarket Type3
8	FDN33	6.305	Regular	0.123365	Snack Foods	95.7436	OUT045	2002	NaN	Tier 2	Supermarket Type1
9	FDA36	5.985	Low Fat	0.005698	Baking Goods	186.8924	OUT017	2007	NaN	Tier 2	Supermarket Type1

	Item_Identifier	Item_Weight	Item_Fat_Content	Item_Visibility	Item_Type	Item_MRP	Outlet_Identifier	Outlet_Establishment_Year	Outlet_Size	Outlet_Location_Type	Outlet_Type
5671	FDA01	15.00	reg	0.054463	Canned	59.5904	OUT049	1999	Medium	Tier 1	Supermarket Type1
5672	NCH42	6.86	Low Fat	0.036594	Household	231.1010	OUT049	1999	Medium	Tier 1	Supermarket Type1
5673	FDF46	7.07	Low Fat	0.094053	Snack Foods	116.0834	OUT018	2009	Medium	Tier 3	Supermarket Type2
5674	DRL35	15.70	Low Fat	0.030704	Hard Drinks	43.2770	OUT046	1997	Small	Tier 1	Supermarket Type1
5675	FDW46	13.00	Regular	0.070411	Snack Foods	63.4484	OUT049	1999	Medium	Tier 1	Supermarket Type1
5676	FDB58	10.50	Regular	0.013496	Snack Foods	141.3154	OUT046	1997	Small	Tier 1	Supermarket Type1
5677	FDD47	7.60	Regular	0.142991	Starchy Foods	169.1448	OUT018	2009	Medium	Tier 3	Supermarket Type2
5678	NCO17	10.00	Low Fat	0.073529	Health and Hygiene	118.7440	OUT045	2002	NaN	Tier 2	Supermarket Type1
5679	FDJ26	15.30	Regular	0.000000	Canned	214.6218	OUT017	2007	NaN	Tier 2	Supermarket Type1
5680	FDU37	9.50	Regular	0.104720	Canned	79.7960	OUT045	2002	NaN	Tier 2	Supermarket Type1

	Item_Weight	Item_Visibility	Item_MRP	Outlet_Establishment_Year
count	4705.000000	5681.000000	5681.000000	5681.000000
mean	12.695633	0.065684	141.023273	1997.828903
std	4.664849	0.051252	61.809091	8.372256
min	4.555000	0.000000	31.990000	1985.000000
25%	8.645000	0.027047	94.412000	1987.000000
50%	12.500000	0.054154	141.415400	1999.000000
75%	16.700000	0.093463	186.026600	2004.000000
max	21.350000	0.323637	266.588400	2009.000000

	Item_Identifier	Item_Weight	Item_Fat_Content	Item_Visibility	Item_Type	Item_MRP	Outlet_Identifier	Outlet_Establishment_Year	Outlet_Size	Outlet_Location_Type	Outlet_Type	Item_Outlet_Sales
0	FDG33	12.857645	Low Fat	0.0	Fruits and Vegetables	172.0422	OUT027	1985.0	Medium	Tier 3	Supermarket Type1	958.752
1	FDW13	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN

	Item_Identifier	Item_Weight	Item_Fat_Content	Item_Visibility	Item_Type	Item_MRP	Outlet_Identifier	Outlet_Establishment_Year	Outlet_Size	Outlet_Location_Type	Outlet_Type	Age
0	FDW58	20.750000	Low Fat	0.007565	Snack Foods	107.8622	OUT049	1999	Medium	Tier 1	Supermarket Type1	24
1	FDW14	8.300000	Regular	0.038428	Dairy	87.3198	OUT017	2007	Medium	Tier 2	Supermarket Type1	16
2	NCN55	14.600000	Low Fat	0.099575	Others	241.7538	OUT010	1998	Medium	Tier 3	Grocery Store	25
3	FDQ58	7.315000	Low Fat	0.015388	Snack Foods	155.0340	OUT017	2007	Medium	Tier 2	Supermarket Type1	16
4	FDY38	12.695633	Regular	0.118599	Dairy	234.2300	OUT027	1985	Medium	Tier 3	Supermarket Type3	38
...	...	...	...	...	...	...	...	...	...	...	...	...
5676	FDB58	10.500000	Regular	0.013496	Snack Foods	141.3154	OUT046	1997	Small	Tier 1	Supermarket Type1	26
5677	FDD47	7.600000	Regular	0.142991	Starchy Foods	169.1448	OUT018	2009	Medium	Tier 3	Supermarket Type2	14
5678	NCO17	10.000000	Low Fat	0.073529	Health and Hygiene	118.7440	OUT045	2002	Medium	Tier 2	Supermarket Type1	21
5679	FDJ26	15.300000	Regular	0.000000	Canned	214.6218	OUT017	2007	Medium	Tier 2	Supermarket Type1	16
5680	FDU37	9.500000	Regular	0.104720	Canned	79.7960	OUT045	2002	Medium	Tier 2	Supermarket Type1	21

	Item_Weight	Item_Visibility	Item_MRP	Item_Outlet_Sales	Age
Item_Weight	1.000000	-0.012049	0.024756	0.011550	0.008301
Item_Visibility	-0.012049	1.000000	-0.001315	-0.128625	0.074834
Item_MRP	0.024756	-0.001315	1.000000	0.567574	-0.005020
Item_Outlet_Sales	0.011550	-0.128625	0.567574	1.000000	0.049135
Age	0.008301	0.074834	-0.005020	0.049135	1.000000

	Item_Weight	Item_Fat_Content	Item_Visibility	Item_Type	Item_MRP	Outlet_Size	Outlet_Location_Type	Outlet_Type	Item_Outlet_Sales	Age
0	9.30	Low Fat	0.016047	Dairy	249.8092	Medium	Tier 1	Supermarket Type1	3735.1380	24
1	5.92	Regular	0.019278	Soft Drinks	48.2692	Medium	Tier 3	Supermarket Type2	443.4228	14
2	17.50	Low Fat	0.016760	Meat	141.6180	Medium	Tier 1	Supermarket Type1	2097.2700	24
3	19.20	Regular	0.000000	Fruits and Vegetables	182.0950	Medium	Tier 3	Grocery Store	732.3800	25
4	8.93	Low Fat	0.000000	Household	53.8614	High	Tier 3	Supermarket Type1	994.7052	36

	features	vif_Factor
0	Item_Weight	4.753126
1	Item_Visibility	2.368765
2	Item_MRP	4.378331

Importing Libraries¶

Context¶

Reading the Dataset¶

Here we read the training Dataset¶

Thus we have 8523 observations(rows) and 12 columns¶

Reading the head upto 10 units¶

Reading the tail upto 10 units¶

Pandas Profiling¶

Describing the dataset and it's information¶

Counting the number of missing values¶

Reading the test Dataset¶

Reading the head of the dataset upto 10 units¶

Reading the tail of the dataset the last 10 units¶

Profiling the dataset¶

Describing the dataset and it's information¶

Counting the number of missing values in the test dataset¶

Entropy Calculation¶

We first calculate the entropy for train Dataset¶

From the entropy calculation above one can say Item_Fat_Content is the most imbalanced variable while Outlet_Location_Type is the most balanced variable in the dataset respectively¶

As it is visible from the graph above Item_Weight is almost uniformly distributed so it's entropy must be high in the training dataset¶

We then calculate Entropy distribution for the testing Dataset¶

From the entropy calculation above one can say Item_Fat_Content is the most imbalanced variable while Outlet_Location_type is the most balanced variable in the dataset respectively¶

As it is visible from the graph above Item_Weight is almost uniformly distributed so it's entropy must be high in the testing dataset¶

Data Cleaning¶

Handling Null Values¶

from box plot Item Weight dosen't appear to have outliers in both datasets, so our mean won't be affected¶

here we can replace the null values with the mode because it is a categorical column¶

EDA(Exploratory Data Analysis)¶

While Low Fat has the majority in the data constituting 64.73% of the products sold Regular constitutes 35.27% of the fat distribution¶

Fruits and Vegetables are largely sold as people tend to use them on daily purpose. Snack Foods too have good sales followed by Household products while Seafood has the lowest Sales¶

Medium Size store has the majority in the data 61.05% in terms of ground covered followed by Small size store and High size store respectively¶

The Outlets are maximum in number in Tier 3 Cities and Tire 3 Outlets comprised 39.1% of the Total number of Outlets while Tire 1 Outlets comprised only 28.02% of the Total Outlets respectively¶

Supermarket Type 1 Outlets comprised maximum 65.43% of the total number of outlets while Supermarket Type 2 Outlets comprised only 10.89% of the total number of outlets respectively¶

Most of the outlets has been running for 38 years since establishment while outlets running for 25 years is comparatively lower than the average outlet count in terms of years running since establishment respectively¶

1)The Item Outles sales are high for both Low Fat and Regular Item types.¶

2)The Outlet Sales is maximum for Medium and High sized Outlets.¶

3)The Outlets we have is Medium and may be with High size Outlets can improve the Outlet Sales.¶

4)The Outlet Sales tend to be high for Tier3 and Tier 2 location types but we have only Tier3 locations maximum Outlets.¶

5)Supermarket Type3 sales tends to be high which in Grocery store is very low¶

6)It is quiet evident that Outlets established 36 years before is having good Sales margin.¶

7)We also have a outlet which was established before 23 years has the lowest sales margin, so established years wouldn't improve the Sales unless the products are sold according to customer's interest¶

While Item Outlet Sales and Item Visibility distribution is positively skewed indicating most of the outlets have sold 2000 items Item MRP and Item weight is more symmetrical while Item weight is more normally distributed respectively¶

1) While light weight Items have a higher amount of Sales overall Item Weight has almost no to very little impact of on Item Outlet Sales respectively¶

2) Item_Visibility has a minimum value of zero. This makes no practical sense because when a product is being sold in a store, the visibility cannot be 0. While Items with lower visibility has a higher amount of Item Outlet Sales Item Visibility has a lower amount of Sales respectively¶

Seafood with Regular fat Content has the highest amount of Sales while Soft Drinks with Regular Fat content has lowest amount of sales respectively.¶

While Type 3 Supermarket in Tire 3 cities has highest amount of sales in Tier 3 and Tier 1 cities Grocery Stores have the lowest amount of Sales respectively.¶

Check if there is any mismatch in the number of features between train and test¶

Data Preprocessing¶

Note other than Outlet Size and Outlet Location Type none of the variables seems to be correlated so problem of multicollinearity seems to be very low¶

As vif of numerical variables is below 5 we might not be concerned about multicollinearity¶

It is always better to check for skewness since data might not be normally distributed¶

As it visibile Item_Outlet_Sales and Item_Visbility is positively skewed hence we apply logarithmic transformation¶

One Hot Encoding the categorical variables¶

To choose the best model we use the Recursive Feature Elimination method and Decision Tree Regression has been used to choose the best features¶

Train Test Split¶

Classification Exercise¶

Linear Regression¶

While the RMSE of our Linear Regression model is 0.5515520555642495 the R2_Score is 71.42586803830434 %¶

Interpretation required¶

Elastic Net¶

Since basic Elastic Model has a bad R^2 score and high RMSE in the presence of default parameters we intend to grid search from the parameters matrix¶

While the RMSE of our Elastic Net Regression model is 0.5513000874440245 the R2_Score is 71.45196938519395 % respectively¶

Interpretation¶

Lasso Regression¶

While the RMSE of our Lasso Regression model is 0.5537688387403791 the R2_Score is 71.19571769469881 % respectively¶

Interpretation¶

Ridge Regression¶

While the RMSE of our Ridge Regression model is 0.5515518531400582 the R2_Score is 71.42586803830434 % respectively¶

Interpretation¶

Decision Tree Regressor Model¶

While the RMSE of our Decision Tree Regression model is 0.7577915929905223 the R2_Score is 0.4606140908409836 respectively¶

Interpretation¶

Random Forest Regressor¶

While the RMSE of our Random model is 0.5678768756082895 the R2_Score is 0.6970936348217647 respectively¶

Ranking¶

Thus inter¶

~ DEVJIT GHOSH¶

~ DEVJIT GHOSH
¶